jiangmuran

jiangmuran / claude-image

Public

教 Claude Code/Codex 真正用好 gpt-image-2 的 skill 包 · Drop-in skill teaching agents to actually use GPT Image 2 — intent-first prompting, surgical edits, parallel batching, visual self-verification

agent-skills ai-skills anthropic chinese-typography claude-code
14
1
69% credibility
Found Apr 26, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A simple installer that adds an image generation and editing skill to Claude AI, allowing users to create pictures from text prompts using a compatible service.

How It Works

1
🔍 Discover the Image Skill

You hear about a handy add-on that lets your Claude AI create pictures from words.

2
📥 Get the Skill Files

Download the simple folder and run the easy setup helper to place it in your Claude folder.

3
🔗 Connect Your Image Service

Enter your account details for the picture-making service so everything can connect smoothly.

4
🧪 Try a Quick Test

Make a tiny sample picture like a red panda to see it work right away.

5
🎨 Create Your Own Images

Tell it what to draw, like 'a cozy cabin in the woods', and watch your picture appear.

6
🤖 Use in Claude Chats

Now Claude spots image requests and uses this skill to make pictures during your conversations.

✨ Images at Your Fingertips

You now have a creative sidekick that turns your ideas into visuals effortlessly.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is claude-image?

Claude-image is a Python drop-in skill for Claude Code and Codex agents, teaching them to properly leverage GPT Image 2 for tasks like generation and editing. It solves the gap where agents fumble image handling by adding intent-first prompting, surgical edits, parallel batching, and visual self-verification. Users get a one-command install that slots into Claude's skills directory, plus a standalone CLI for generating images from prompts or editing existing ones via an OpenAI-compatible API.

Why is it gaining traction?

It stands out by tuning Claude Code Codex agents specifically for gpt-image-2 at a custom endpoint, with CLI features like parallel requests for faster batching, high-quality defaults at no extra cost, and strict validation for sizes up to 3840px. Developers notice the seamless integration—no deps, idempotent setup, even optional credential handling and smoke tests. Early benchmarks show it boosts Claude image analysis reliability over generic prompting.

Who should use this?

AI agent builders tweaking Claude Code Codex subagents or MCP setups for visual tasks, like claude github actions generating diagrams or claude github copilot plugins needing image edits. Frontend devs prototyping UI mockups via CLI, or teams running claude code codex benchmarks on multimodal workflows.

Verdict

Worth a quick install for Claude Code Codex skill experiments—solid CLI and agent teaching at 14 stars, though the 0.7% credibility score flags it as early alpha with basic docs. Test on non-critical projects first.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.